Investigation Into Production of Refined Bismuth

7/21/2019 Investigation Into Production of Refined Bismuth

1/18

COMMONWEALTH OF AUSTRALIA

D E P A R T M E N T O F N A T I O N A L D E V E L O P M E N T

BU R E A U O F M IN E R A L R E SO U R C E S

G E O L O G Y A N D G E O P H Y SIC S

RECORDS:

1965/182

INVESTIGATION INTO PRODUCTION OF REFINED BISMUTH

(A.M.D.L. Report No.388)

by

R.E. Wilmshurst

The information contained in this report has been obtained by the

Department of National Development, as part of the policy of the Common-

wealth Government, to assist in the exploration and development of mineral

resources. It may not be published in any form or used in a company

prospectus without the permission in writing of the Director, Bureau of

Mineral Resources Geology and Geophysics.


2/18


A . M . D . L . R e p o r t N o . 3 8 8 )

by

R.E. Wilmshurst

Records 1965/182

T a b l e o f C o n t e n t s

. Introduction

A.M.D.L. Report No.388

page 1

Title Page

Contents

pages 1-12

Figure 1

The information contain ed in this report has been obtained

by the Department of National D evelopment, as part of the policy

of

the Commonwealth Government to assist in the exploration and

development of mineral resources, It may not be published in any

form or used in a company prospectus without the permission in

writing of the Director, Bureau of Mineral Resources, Geology and

Geophysicso

of the Australian Mineral Development Laboratories, Adelaide.


3/18

Introduction

Marketing of crude bismuth metal produced by

leaching ore from the Tenna nt Creek area proved difficult.

The relatively low return for crude bismuth compared with

prices listed for pharmacev. Aical grad e material prompted the

Bureau to sponsor this research into an electro,-refining

process suitable for the production of bismuth metal of

acceptable _grade at Tennant Creek.

The report has been included in the Bureau's Recol

d

series so that it may enjoy the wider distribution afforded.

by the open file system. It has not been altered in any

w a y .


4/18

2/1/4

MDL Report 388

November,

1964


by

R. E. Wilmshurst

to

THE BUREAU OF MINERAL RESOURCES

Investigation by: Industrial Chemistry Section

Officer in Charge: F. R. Hartley

L. Wallace Coffer, Director

THE AUSTRALIAN MINERAL DEVELOPMENT LABORATORIES

Adelaide

outh Australia


5/18

CONTENTS

Page

1

INTRODUCTION

1 1

lectro-refining of Bismuth

2

SUMMARY

3

MATERIAL EXAMINED

4

EQUIPMENT

5

EXPERIMENTAL PROCEDURE AND RESULTS

5. 1

reliminary

5. 2

aboratory Investigations

5. 2. 1

mprovement of Existing Process

5. 2. 2

efining of Ingot Metal

5.3

iscussion

6

PILOT SCALE REFINING OF BISMUTH METAL

7

DISCUSSION

1

8 REFERENCES

2

FIGURE 1


6/18

1.

NTRODUCTION

Early in 1962, a request was received from Mr C. N. Strong of

Tennant Creek for the development of an electro-refining process to produce

pharmacutical grade bismuth from crude metal. The latter was produced

from tailings dumps in the Tennant Creek area by a process similar to that

described by Cathro and Koch l

(a). In his process, Mr. Strong leached the

ore by percolation with sodium chloride-sulphuric acid solution. The leach

solution, together with liquor used to wash the leach residue, was then

treated with scrap iron. Bismuth metal in a finely divided sponge form

was precipitated in this cementation step, and was subsequently melted into

ingots. The product, although technically a good quality metal, was not of

sufficiently high purity for use in the manufacture of pharmaceuticals.

In September, 1962, Mr Strong approached the Commonwealth

Bureau of Mineral Resources for support in financing the investigation.

This support was given and work on the project commenced in January, 1963.

Specifications for pharmaceutical bismuth are laid down in the

British Pharmaceutical Codex, but this authority uses arbitrary tests

extensively, without giving numerical limits for specific impurities. Tests

of this nature are convenient as acceptance tests to determine the suitability

of any given material, but are inadequate for use in investigational work.

Evaluation of the tests and correspondence with Bismuth Products Pty. Ltd.

led to the following tentative limits for impurities in pharmaceutical grade

bismuth metal:

Arsenic not more than 2

pm

Lead

itto - 0.04 %

Copper

0 ppm

Silver

.01

Antimony

0

pm

Tellurium

pm

The experimental work described in this report was aimed at

production of a metal to this specification by an electro-refining process

suitable for use at Tennant Creek.

1.1

lectro-refining of Bismuth

Electro-refining of bismuth has been practised for many years but

little detailed information has been published. Gruzensky and Crawford

z

surveyed the various electrolytes recommended from time to time and found

only two to be satisfactory. One of these, a hydroflusilicic acid solution,

gave excellent results but was unstable and unsuitable for large scale use,

particularly in remote areas. The alternative hydrochloric acid-bismuth

chloride solution gave satisfactory results except that removal, of lead was

not achieved; this was rectified by the addition of sulphuric acid. Kern

and Jones

3

recommended a similar solution, and this was used also by Cathro

and Koch (loc cit) in refining of Tennant Creek bismuth.

(a) See Section 8 for References.


7/18

2

Metal from Tennant Creek has been variable in composition, with

analyses generally in the following ranges:

Copper

.03 - 0.20 %

Lead

.05 - 0.10 %

Zinc

a) < 0.01

Silver

50

pm

Iron

0.01 %

The work of Cathro and Koch, in conjunction with consideration of

standard electrode potentials, indicates that simple electro-refining applied

to this metal would have the effect shown in Table 1.

TABLE 1: LIKELY PURIFICATION OF BISMUTH

Element

Crude Metal

Reduction Factor

Refined Metal

Cu

0.2

5

0.04

Pb

0.1

450

2

Pm

Zn

0.01

high

low

Fe

50

pm

high

low

Ag

0.01

1.3

70

Pm

These figures suggest that electro-refining alone is unlikely to

produce a metal to the required specification. Pyrometallurgical methods

of removing these impurities were therefore investigated. Some considera-

tion was also given to modifying the existing operations at Tennant Creek to

improve the crude metal purity.

2.

UMMARY

In response to a request from Mr C. N. Strong of Tennant Creek and

under the sponsorship of the Commonwealth Bureau of Mineral Resources,

an electro-refining process for metallic bismuth has been developed.

Simple electro-refining using an acidic chloride electrolyte was incapable of

producing pharmaceutical grade bismuth from crude metal. Various

pyrometallurgical techniques were examined and a hybrid process developed.

Melting of the crude metal with sodium hydroxide and sodium cyanide,

casting the resultant metal into anodes, and electro-refining in the chloride

bath, followed by a sulphur drossing step gave pharmaceutical quality

bismuth metal.

(a)

Less than.


8/18

3

A large parcel of bismuth metal has been refined by this process to

give over 200 pounds of pharmaceutical quality metal. The process is

capable of good yields of metal and the electro-refining operation has a high

current efficiency.

The weaknesses inherent in electro-refining processes are considered

to make them adaptable only with difficulty for small-scale use in remote

areas such as Tennant Creek.

3.

ATERIAL EXAMINED

The ore used in leaching experiments was from the Jubilee Mine,

near Tennant Creek. This material contained approximately 3 per cent

bismuth.

In both the small scale and pilot scale refining experiments, a

parcel of 350 pounds of crude bismuth metal from Mr Strong was used. This

parcel consisted of 14 ingots each of 25 pounds nominal weight. As pre-

liminary experiments had indicated that copper and silver were the elements

most likely to cause difficulty in refining, each ingot was assayed for these

metals. The results of these assays are set out in Table 2.

TABLE 2: ANALYSES OF CRUDE BISMUTH INGOTS

Ingot

No.

Silver

ppm

Copper

23(a)

0.074(b)

2

16

0.076

3

12

0.113

4

18

0.062

5

8

0.099

6

10

0.063

7

18

0.035

8

20

0.070

9

18

0.110

10

18

0.098

11

18

0.159

12 18

0.160

13

14

0.119

14

16

0.206

a)

Mean, values from 22 ppm to 26 ppm in various

parts of ingot.

b)

Mean, values from 0.051 to 0.093 .


9/18

4

4.

EQUIPMENT

The pilot scale electro-refining experiments were conducted in a

multiple cell bank constructed for the purpose. This vessel was 3 feet

square in plan and 18 inches deep, and was divided by vertical partitions into

ten cells 18 inches by 7 inches in plan and 18 inches deep. Welded steel

construction was used and all internal surfaces of the tank were covered by

epoxy resin reinforced with glass fibre.

Two copper cathodes, approximately one sixteenth of an inch thick,

were suspended in each cell compartment with an area 15 inches square

immersed in the electrolyte. Only five cells were used and these were

connected electrically in series. The current for electrolysis was obtained

from a silicon diode rectifier set rated to deliver 50 amperes at 2 volts DC

from a 240 volt 50 cycle AC input. A variable auto-transformer in the AC

circuit served to control the rectifier output.

5 .

EXPERIMENTAL PROCEDURE AND RESULTS

5.1

reliminary

An experimental pyrometallurgical treatment was begun at Tennant

Creek by Mr Strong. The procedure is given as described by him in

Figure 1.

The metal corresponding to Sample B was cast into an anode and

electro-refined. Sample C was the cathode deposit.

Analyses by AMDL of these samples, for copper and silver only,

are given in Table 3.

TABLE 3: ANALYSIS OF BISMUTH METAL

Sample

opper

ilver

PPm

Pm

A

60

0

30

0

15

0

3300

0

170

0

490

0

55

0

65

0

110

27


10/18

5

Samples G and H, also from Tennant Creek, are respectively

before and after a cyanide leaching treatment of sponge metal. Samples J

and K are from a similar experiment conducted at ANIDL.

In general the following conclusions can be drawn:

a.

the copper tends to concentrate in the skimmings

(Sample D);

b.

the sulphur and cyanide flux treatment reduces

copper appreciably (Samples A-B and D-E);

c.

electro-refining following the flux treatment

produces a metal of purity approaching the required

level;

d.

copper can probably be removed also by cyanide

leaching of the cemented bismuth sponge before

smelting.

5.2

aboratory Investigations

5.2.1

mprovement of Existing Process

To simulate the existing leaching process, 250 grams of the ore

sample submitted by Mr Strong was crushed to pass a

/-inch screen, and

then leached twice by percolation with the same 500 ml of leach solution.

This solution contained 10 per cent w/v NaC1 and 4 per cent w/v H

z

SO

4

. A

further 100 ml of fresh solution was used as a wash.

The leach liquor was found by analysis to contain 12.1 grams of

bismuth per litre. It was divided into three equal parts. Part 1 was

cemented for 24 hours with sheet steel. Part 2 was cemented for 2 hours

with 1 gram of bismuth sponge, filtered and then cemented with iron as in

Part 1. Part 3 was treated with 0.1 gram of sodium sulphide, stirred for

2 hours, filtered and cemented with iron as in Part 1.

Cementation with metallic bismuth was expected to remove copper

from the solution, while the partial sulphide treatment of Part 3 should

precipitate copper and some bismuth as sulphide. As the amount of metal

in each experiment was quite small, the bismuth sponge products were

oxidised to Bi

2

03before analysis.

The results are as follows:

Cu

Ag

Part 1:

Control only 61


11/18

6

5.2.2

efining of Ingot Metal

The metal used in the experiments was from an ingot previously

sampled and found to contain 16 ppm of silver and 0.206 per cent copper.

This metal was treated by a number of different techniques with the following

results.

Simple Electro-refining. A sample of the metal was cast into a sheet 8 inches

square and approximately 1/4-inch thick. This was electro-refined on to a

waxed copper sheet at a current density of 10 amperes per square foot. The

electrolyte contained 50 g per litre HC1, 70 g per litre bismuth (added as

BiC1 3

) and 0.1 g per litre pyrogallol. The electrodes were 11/4 inches apart,

and the solution was agitated gently. Operating voltage was 0.13 volt.

The cathode deposit was granular, with a tendency to growth in fingers, but

adequately adherent. The analyses of this metal indicated 61 ppm copper

and 3 ppm silver.

This experiment confirmed that simple refining was not adequate.

Use of Parkes Process. One hundred grams of metal and 10 grams of

powdered reagent grade zinc were melted and held at 450

C for 2 hours with

occasional stirring. After settling and solidification, the top half of the

specimen was rejected and the bottom half sampled. A control test without

zinc was carried out simultaneously.

The analyses were:

g

pan pprr_

Control

1

Experiment

The results suggest that the zinc treatment was successful in

removing copper and silver to acceptable limits but, more important, that

careful settling without zinc was also effective.

Use of Sulphur Drossing. A sample of metal was fused with 1 per cent of

its weight of elemental sulphur, using a salt bath for ease of temperature

control. The melt was skimmed and cast into an anode. The electro-

refining process described above was carried out and gave a metal containing

4 ppm copper and 3 ppm silver.

This technique therefore gave a metal acceptable with regard to

both copper and silver.

Cyanide Fluxing. A sample of metal was fused with 1 per cent of its weight

of sodium cyanide. Some difficulty was experienced and contact was not

ideal. The treated metal was cast into an anode and refined. Two products

were collected - one with the electrolyte described above, and the other

after addition of 75 g per litre of H

2

SO

4

to the solution. The products

contained:


12/18

7

Product

With Sodium Cyanide

ith Sodium Cyanide plus

H 2 S O 4

Copper,

Cu

15

ppm

Copper,

Cu

19

PPm

Silver,

Ag

3

a)

ppm Silver,

Ag

3

PPm

Lead,

Pb

0.015 Lead,

Pb

0.015

The cyanide treatment effectively reduced the copper and silver

concentrations. The addition of sulphuric acid to the electrolyte made no

difference to the concentration of lead in the products.

5.3

iscussion

The experiments above indicated that simple electro-refining of the

bismuth ingot metal was not likely to produce a metal to the required

specification.

reliminary treatment with sulphur, sodium cyanide or

metallic zinc, followed by electro-refining, should produce an acceptable

metal. In the latter case the refining would be necessary to remove zinc

left in solution in the bismuth. Of these methods, the zinc treatment would

probably give a lower yield than the others, because of difficulty in separating

the two metals. The sulphide treatment would give a greater loss of bismuth

in the skimmings than the cyanide method, and the latter is therefore

preferred.

The removal of copper by careful settling was not fully understood,

and may have been due partly to segregation on cooling. However, it was

observed in the experimental work that some sulphur was present in the

ingot material - a sample of skimmings from an ingot was found to contain

0.2 per cent of sulphur, presumably as copper and bismuth sulphides. This

probably accounted for copper removal by settling. Use of the method for

refining is feasible but bismuth losses would be high.

The presence of the sulphide was noticeable also in the electro-

refining process. A black crystalline deposit, identified as bismuth

sulphide, appeared at times on the anode and in the anode sludge. This

phenomenon was also associated in some way with the appearance of a white

coating of bismuth oxychloride on the anode in some experiments. This

coating, if left unchecked, blanketed the anode surface and increased the cell

voltage to 0.25 volt or higher. Both of these effects are probably associated

with gradual loss of free acid from the electrolyte.

At this point a complete refining process could be formulated.

To check the validity of this proposed process, a separate proving

experiment was carried out.

A sample of ingot metal was melted in a steel vessel using as flux

a mixture of equal weights of sodium hydroxide and sodium cyanide. The

total weight of flux was 2 per cent of the weight of bismuth. The mixture

was held at 400-500

C for 10 minutes with occasional stirring, and then

allowed to cool slowly. As solidification commenced about 2 per cent of the

bismuth was skimmed off using a steel gauze and discarded. The bulk of

the metal was reheated and cast into anodes.

(a)

ess than.


13/18

8

The anode metal was electro-refined using a hydrochloric acid

electrolyte containing 70 grams of bismuth, 50 grams of free acid, and

0.1 gram pyrogallol per litre. Copper cathodes were used and were coated

with a thin film of paraffin wax to assist in removal of the deposit. The

current density was 10 amperes per square foot, and the electrolyte was

agitated gently. No diaphragm or anode bag was used.

Over a period of 90 hours, 2020 grams of metal was deposited at a

current efficiency of 94 per cent. In the following 72 hours, 1620 grams of

metal was deposited at virtually 100 per cent efficiency. Analyses of the

two deposits, designated A and B respectively, are shown in Table 4 below.

TABLE 4: ANALYSES OF BISMUTH METAL

4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 = 1 = 1 1 1 1 7 1 1 1 1

Element

Deposit

A

PPm

Deposit B

PPm

Limit

PPm

Lead


14/18

9

During the whole of this period, metal was deposited on the cathodes

at high current efficiencies, except in some cases where the deposit on the

cathodes bridged the space between the two electrodes. When this happened

the cell was effectively short circuited and refining ceased. In an effort to

prevent this difficulty glycerol was tried as an alternative to the pyrogallol

used to modify the character of the deposit throughout the earlier experimen-

tal work. Little if any improvement was evident. On some occasions the

weight of metal deposited on the cathodes before the cell became unworkable

was quite small.

Removal of the deposit from the cathodes was not difficult but always

involved some chance of contaminating the product with cathode material.

The risk was proportionately greater when the cathode deposit was thin and

refining had been interrupted by bridging of the electrodes.

No purification of the electrolyte was attempted at any stage, but

additions of hydrochloric acid were required at intervals to replace acid

lost by evaporation and by side reactions in the cells. At the completion of

the run the cathode deposits were melted down using sodium hydroxide flux

and cast into ingots which were sampled for analysis. Results of the

analyses are shown in Table 5.

TABLE 5: ANALYSIS OF REFINED BISMUTH

Expressed in parts per million

Ingot Number

Ag

Cu As

Te

Pb

Sb

5 76 2

40

10

2

9 69

2

18

10

3 5

140

2

35

10

4 6

99

2

15 10

5

3

37

2

10 10

6 2

3 7

2

15

10

7

7

755

2

8

10

8

4

110

2

18

10

9 3

170

2

15 10

10 6

125

2

8

10

11

6

150

2

10

10

These analyses indicate that the refined metal was satisfactory in

all respects except that it contains considerably more copper than is permissible.

Following the earlier work it was thought that this might be removable by a

skimming technique. Accordingly all of the refined metal was re-melted

using caustic soda flux, and small amounts of skimmings were taken as the

metal solidified. These skimmings were kept separate, the bulk of the metal

being re-heated and cast again into ingots. Results of the analysis of these

ingots for copper are given in Table 6.


15/18

10

TABLE 6: ANALYSIS OF SKIMMED BISMUTH

Expressed in parts per million

Ingot Number

Cu

1070

2 320

3 290

4

440

5

750

6

500

7

460

8

260

9

670

Results of this operation were disappointing in that the amount of

copper had increased rather than decreased. The increase may have been

due in part to sampling errors but was almost certainly caused largely by

contamination from an unidentified source. Clearly the metal was still

unsatisfactory.

At this stage a sulphur drossing technique was used to reduce the

copper content of the bismuth. The ingot material was melted once again

using sodium hydroxide flux and, while molten, the metal was treated with

elemental sulphur. The weight of sulphurused was 1 per cent of the weight

of the metal in any given melt and it was added slowly with stirring to the

bismuth. The molten metal was then allowed to cool, was skimmed, re-

heated, and recast into ingots. Results of further analyses are given in

Table 7.

The amount of copper present was then well within the specified

limit with the exception possibly of the ingot designated 9A. However, other

ingots were so far below the limit that the metal could all be considered

satisfactory after a single blending operation.

The weights of finished refined metal, skimmings, and crude

residual metal were respectively 214, 6 and 34 pounds. While the overall

yield could be regarded as poor, considerable quantities of bismuth were

discarded at various stages of the operation because it was considered that

recovery would cost appreciably more than the value of the recovered metal.

In practice excellent yields of metal should be possible.


16/18

11

TABLE 7: ANALYSIS OF DROSSED METAL

Ingot Number

u

eight of Ingot

PPm

b

lA 5

24

2A

16

24

3 A

8

20

4A

20

21

5 A

17

19

6A

17

21

7A

22 20

8 A

15

201/2

9 A

5 9

221/2

10A

20

221/2

7.

ISCUSSION

The investigation was restricted in scope as the original request

was for development of an electro-refining process. The experimental work

demonstrated clearly that a simple electro-refining process was not capable

of producing metal of the required purity. Consequently a combination

process was developed in which pyrometallurgical steps were combined with

electro-refining to produce pharmaceutical grade bismuth. The overall

process has been used to refine substantial quantities of metal and has been

shown to be effective.

In spite of the reasonable results which were obtained from the

refining process, electro-refining can scarcely be considered a satisfactory

refining process in the present context. Its inability to remove certain

impurities and the possibility of impurity build-up in the electrolyte with

subsequent falling-off in efficiency of refining are strong arguments against

a process of this type. Control of the nature of the cathode deposits by use

of various additives to the electrolyte is an art rather than a science. With

experience a satisfactory technique could probably be developed. However,

there seems to be an inherent weakness in any process which involves

plating of metal on to a cathode of material which would be an undesirable

contaminant in the product. Removal of the deposits from the copper

cathodes was relatively easy but when the copper specification is so stringent

the danger of contamination beyond the specified limit is always present.

Possibly this could be averted by use of an alternative cathode material.

With further development, use of bismuth cathodes may be feasible but this

has not been attempted in the present investigation. However, it is consid-

ered that a pyrometallurgical process along conventional lines using such

operations as sulphur drossing, skimming, partial oxidation, chlorination


17/18

12

and so on, could give metal of an acceptable quality rather more simply than

the refining processes described above. Nevertheless it has been shown

that a dombination process employing sulphur drossing, skimming and electro-

refining can produce pharmaceutical grade bismuth from crude metal

produced in the Tennant Creek area.

8.

EFERENCES

1.

CATHRO, K. J., and KOCH, D. F. A., Recovery of Bismuth from

Chalcopyrite Calcines , paper presented at the A. I. M. M.

Hydrometallurgy Symposium, Adelaide, February, 1960.

See also Aust. J. App. Sci. 13, (3) 175 (1962).

2. GRUZENSKY, P. M., and CRAWFORD, W. J., U. S. Bureau of

Mines, Report of Investigation 5182.

3.

ERN, E. F., and JONES, T. R., Trans. Amer. Electrochem.

Soc. 57, 255 (1930).


18/18

FIGURE

1

EXPERIMENT L

PYR ONIE:r Al L LIR GIG AL

ReprrmETN

Crude Metal

Skim at 600

Metal

Ximmings

Sample

A

ample I:1

Flux with Sulphur

LM

with ulphwr

Cy

anide

I

i

Metal

5arrtpla

E

Flux with 7 inic

aizeP

hen Sulphur and Cyan ide

1

Flux Discarded

etal

Sa.rnple

F

and Cyanide

and

Metal

Flux Flux

Sa mole

1 3

b rd

Disc an:led

Z G

PR e_h

Documents

Investigation Into Production of Refined Bismuth